Guides / Guide
An order book is the list of resting offers to buy and sell at each price. A prediction market book works the same mechanically, but what it quotes is different in a way that changes how you read it: the price is a probability, and it is bounded at 0 and 1.
A binary contract pays $1 if the outcome happens and $0 if it does not. So a resting bid at 34¢ is someone saying "I'll pay 34¢ for a dollar contingent on this" — that is, they price the outcome at roughly 34% likely. The best bid and best ask bracket the market's live implied probability, and the mid is the usual summary of it.
Two consequences follow from the 0–1 bound. Prices cannot run away, so the book is range-bound in a way an equity book is not. And a fixed spread means very different things at different levels: 2¢ wide at 50¢ is about 4% relative uncertainty, while 2¢ wide at 5¢ is 40% — which is why long-shot markets look permanently illiquid when measured in cents.
On Kalshi, the book is quoted as Yes and No. These are not separate markets: a
No bid at 40¢ is economically a Yes offer at 60¢, because the two outcomes must
sum to $1. This is why ProbSights' Kalshi snapshots carry yes_bids and
no_bids — they are the two halves of one book — while also exposing normalized
best_yes_bid, best_yes_ask, mid, and spread
so you don't have to do that conversion for a top-of-book read.
Polymarket quotes bids and asks against a specific outcome token
(asset_id), which reads more like a conventional book. Its crypto series are framed
as Up/Down over a fixed window rather than Yes/No on an event, but the structure is identical.
Each side is a list of { price, size } levels. What matters is not the top price
but how much size sits behind it:
A last-traded price tells you where one trade happened. The book tells you whether you could trade there, in your size, right now. For a market with a 3¢ spread and $2,000 of depth, those are wildly different claims — and assuming the mid is exactly the error behind most over-optimistic backtests.
ProbSights records these books as timestamped snapshots, so you can replay how depth and pricing evolved toward resolution — see Polymarket and Kalshi historical data, or the historical data tool to look at one market interactively.
Data from Kalshi and Polymarket. Search live API docs Pricing